Investigation into the Physical Properties Responsible for the Formation of Basaltic Spatter

Abstract:

Pyroclastic material from basaltic eruptions takes many forms, including ash, scoria, spatter, and clastigenic flows. These products define a spectrum that is a function of clast temperature, how clasts interact with each other when deposited, and the extent of welding or fusing as clasts cool. The pyroclastic deposit properties depend largely on microscopic structures developed at clast interfaces: a) scoria clasts retain their original shapes after deposition; b) agglutinated clasts deform but have a limited weld; c) stronger welding occurs when neighboring clasts deform and their contacts are fused; and d) fusion results when clasts have completely merged with one other. We examine conditions necessary to produce spatter using 2 experimental approaches: small-scale tests in a muffle furnace, which vary the parameters of clast size, deposit thickness, and temperature; and large-scale tests at Syracuse University’s Lava Project facility. Starting material is Keweenaw Peninsula basalt (~50 wt.% SiO₂, 7.5 wt.% FeO, 7.3 wt.% MgO). In the furnace, pea-sized chips (0.5-7 mm) begin to deform at 1100^oC, and agglutination occurs from 1100 to 1150^oC, with clasts beginning to weld ~1125^oC. Clasts are completely molten at ~1150^oC. Furthermore, the thicker the deposit, the more extensive the welding in the middle, owing to greater heat retention by overlying material. To examine the role of clast size, crushed samples were sieved into 3 fractions and heated; the smallest clasts with the most surface area and least thermal mass were most extensively welded. Finally, temperature experiments reveal that at 1100^oC, boundaries between clasts are 29% welded (extent of welding is measured by comparing the length of welded contacts between clasts with that of unwelded contacts in SEM images). At 1125^oC, clasts are 62% welded, and at 1150^oC, they are fully welded. Above 1150^oC, clast shapes are no longer discernable. These results are consistent with work by Rader (2014), conducted at 1130^oC on basalts from Devil’s Garden, OR. Our results indicate that a potentially wide window exists for spatter formation, depending on eruptive temperature and cooling rates upon deposition. We will also report results from large-scale simulations performed at the Syracuse University facility in late August.